Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[woodyl]

I have an X10 setup which includes a few lamp modules, 3 wall switch modules, an appliance module, 3 DS11A sensors, and a chime module.  Some of the modules are soft-start and some are old-style start.  My light bulbs are all incandescent.

I have a CM11A,  a CM19A, and a CM15A in use.  The CM15A is associated with ActiveHome Pro and the other 2 control modules use Linux with Heyu or mochad and custom apps.

I also have an RR01 repeater.

Everything is set up on House Code E except the chime module which uses House Code K.

Recently, I started experiencing regular firestorm events.  These seem to include a series of All_Lights_On commands then a few All_Lights commands.  The whole sequence will run for about 2-3 minutes, then stop.  This seems to happen at various times.  Recently, it happened around 2AM, 5AM and 6AM.  It does not seem to happen much in the middle of the day.

The commands only seem to affect the House Code E devices; the chime module on House Code K never sounds.

At I thought that one of my command modules was causing the problem, so I started monitoring them to see what was happening.  From monitoring, it appeared that the signals were getting sent out on the powerline, and not via RF.  I systematically disconnected everything that was capable of sending a signal, i.e., the command modules, and the repeater.  Even with all sending devices removed, I am still getting the firestorm.

I have a smart meter, but it was installed in 2009, and everything has worked fine for years, so I don't think that's the problem.

The thing that's strange to me is that the signals only seem to be affecting the E code and never the K code.  That makes me think that it's not random signal noise.  That, plus the fact that it's happening every night makes it seem like something other than random noise. The next step for me would be to change my House Code, but before I do that, I want to see if I can figure out the source.

At this point, I would appreciate any help in diagnosing this.

[Brian H]

RR01 repeater? Are you maybe thinking of an RR501 Transceiver that is not a phase coupler/repeater?
There are some phase coupler repeater that have been known to cause firestorms with a CM15A.
http://kbase.x10.com/wiki/Known_Issues_with_ActiveHome_Pro

You may also want to unplug the RR501 {if that is what you have} and see if things get better. I had one start to fail and all kinds of strange things where happening in my system. I also have a TM751 that sends dim firestorms and is now in the electronic scrap box.

[dhouston]

I also have a TM751 that sends dim firestorms and is now in the electronic scrap box.

Not to hijack the thread but you might have fixed the TM751 by moving it or reorienting the antenna. It was inductively coupling to the 120kHz on the powerline - I could start/stop the storms at will by aligning the antenna with/across the power cable of the powerstrip it was in. Of course, it would not have happened if X10'S programmers were a bit more professional.

On topic, I believe their was a similar problem with the CM11A causing signal storms with a Leviton repeater.

[woodyl]

Actually it is an RR501 repeater, but as I said, I've disconnected the repeater and the control modules.  I just discovered, though, that the CM11A was disconnected from the computer, but not the wall.  I've disconnected that, too, so I'll see if the firestorm stops.  At this point, I believe that every device capable of sending a signal is disconnected (except for the DS11A security modules and they are RF only and use frequencies that are distinct from regular switch modules).

[Brian H]

Thanks for the clarification on the RR501.
To most of us a RR501 is a transceiver. Takes the X10 RF commands and sends them on the power lines.
A Repeater is a hard wired module between the homes split phase Power Lines to resend X10 power line commands on the other phase of the wiring.

[dhouston]

I just discovered, though, that the CM11A was disconnected from the computer, but not the wall.

If disconnected from the computer with the serial cable still connected to the CM11A, that's the likely source of the storms.

[woodyl]

I just discovered, though, that the CM11A was disconnected from the computer, but not the wall.

If disconnected from the computer with the serial cable still connected to the CM11A, that's the likely source of the storms.

That's my best guess, so far.  Are you saying that a CM11A disconnected from a computer is more likely to create a firestorm than it would be if connected to the computer?  In my case, it appears that the firestorm happened before and after I disconnected the cable from the computer (but while the CM11A was still plugged into the wall).

Otherwise, has anybody heard of a CM11A sending out rogue signals like this before?  I was using it without batteries, and without any commands stored in the controller, so I would not expect it to send any signals other than the ones initiated on the PC.  Of course it's possible that something got stored in the CM11A unintentionally or else that the controller went bad in some way.

[Brian H]

There is some information on odd things occurring with the CM11A on this web site. I know lockups and overheating where reported. I didn't see rouge transmissions but it could be.
http://www.idobartana.com/hakb/index.htm
Part way down the left pane of the web page.
I believe it was originally from Dave Houston's information.

[dhouston]

I was using it without batteries, and without any commands stored in the controller, so I would not expect it to send any signals other than the ones initiated on the PC.

When plugged into power the CM11A sends signals out the serial port whenever there is X10 activity on the powerline. When the serial cable is dangling from the CM11A with no GND connection to the PC, these signals get capacitively coupled back to the CM11A serial input creating weird results.

[woodyl]

I was using it without batteries, and without any commands stored in the controller, so I would not expect it to send any signals other than the ones initiated on the PC.

When plugged into power the CM11A sends signals out the serial port whenever there is X10 activity on the powerline. When the serial cable is dangling from the CM11A with no GND connection to the PC, these signals get capacitively coupled back to the CM11A serial input creating weird results.

Probably doesn't help that I'm also using a USB-to-serial device to connect this to my Linux PC.  We'll see if the firestorm goes away now that the CM11A is unplugged from the wall.

[dhouston]

I'm not sure whether the USB-serial adapter makes much difference. Most of the reports go back several years before PCs had USB ports. I suspect the modular cable used for the CM11A serial port was a factor.

The fact that you have CM15A, CM11A & RR501 might be a contributing factor. All are polite in that they will stop transmitting if they sense a collision (see a 1-bit when they send a 0-bit) and all then will try to retransmit once the line is clear (6 quiet half-cycles of the 60Hz) but they will each wait a pseudo-random number of half-cycles before trying again. The problem is that when there are multiple polite transmitters it's almost guaranteed there will be new collisions. 10-12 years ago I ran a lot of tests forcing collisions and trying to work out the details without much success (the results were very confused and confusing) but I do recall instances where my RR501 would appear to go berserk and transmit for very long periods.

I had extensive notes and data from all the testing I did years ago but, when I retired an old Windows 95 PC, I copied my X10 directory to a network HDD. Unfortunately, the network drive failed a few years back and I lost it all. While I still have the old PC in storage, my health is too far gone for me to try to get it going again.

[woodyl]

Just to follow up, it appears that the problem was caused by the CM11A.  I've turned on the other devices (CM15A, CM19A, RR501) and they all seem to be playing together well, now.  The CM11A was not supposed to be sending any commands at the time the storms happened, so my guess is that something was corrupted and stored in the device.   

Dave: thanks for the info on collisions when using 2 controllers.  In my case, the CM15A is sending powerline only and the CM19A is sending RF only, so it doesn't see commands sent by the CM15A.  Also, the CM19A is mostly used to monitor DS12A sensors and when it controls other devices, they're on a different house code.  I'll keep the collision situation in mind if that should happen, though.